采用直接体绘制方法显示医学数据时,一般通过修改传递函数来显示感兴趣区域,但仅通过调节传递函数很难清晰地显示被遮挡的人体组织,为此提出一种新的分层剥离直接体绘制算法.通过计算相邻的不透明度极小值和极大值之间的斜率,并根据用户指定的斜率阈值和累积不透明度阈值确定分层点,实现对体数据的分层绘制.实验结果证明,该算法可以根据不同医学数据的特征信息更精确地定位分层点,绘制效果清晰,速度较快.
To visualize medical volume dataset, we often render the regions of interest with an appropriate transfer function. However, the occluded regions are unable to be clearly displayed by merely modifying the transfer functions. In this paper we present a novel layered peeling algorithm to direct volume rendering for medical dataset. By the algorithm, the opacity slope in terms of the minimum and maximum values of adjacent opacity, is calculated, and the layers as defined by the threshold of the slope given by the user and the accumulated opacity, are used to render the volume dataset. Our approach can be used to accurately locate the layer boundaries associated with the features of medical volume dataset, and experiments show that our method can be employed to produce better quality images at a high speed than previous methods.